Suppose that we are interested in estimating the effect that companies’ spendings on R&D have on their profitability. For this, we estimate the regression model where profit-to-sales ratio is our dependent variable and the R&D spendngs-to-sales ratio is the main explanatory variable.

We also include in the regression the dummy variables that are indicators of the sector where company works. Let’s say that there are five sectors and each company belongs to one (and only one) of them. Then we include five dummy variables in the regression. Each of them takes value 1 if the company belongs to this sector and 0 otherwise.

Our regression command in Stata is: regress prof rd_spend d1 d2 d3 d4 d5
where prof is profit-to-sales ratio variable, rd_spend is R&D spendngs-to-sales ratio variable, d1-d5 are dummy variables for the sectors.

1) When we run this regression in Stata, it drops one of the industry indicator dummy
variables. Why is that?
2) How do we interpret the size of the regression coefficient at R&D spending variable? (Provide ceteris paribus interpretation)
3) How do we interpret the size of the regression coefficient at any industry dummy?
4) Can we claim that we estimated a causal effect of R&D spending on profitability? Why or why not?